Test your Network
Load, run, get results and compare your network in different environments.
Documentation
More information about the project. References used, guides, articles, etc.
We are always open for new contributions from outside developers. If you want to contribute to this project, please visit the CONTRIBUTING guide
Behavior Metrics is a tool written in Python that, using the JdeRobot environment, allows to compare different autonomous driving networks as well as programs made manually.
This project aims to have a platform for evaluating and testing complex behaviors for different robots using machine learning and deep learning algorithms. This application provides different functionalities such as:
- Loading a simulated environment for different scenarios where you can evaluate complex robots behaviors.
- Generating datasets to train your models to be tested later.
- Evaluate the performance of your models by comparing them against other models.
- Change the scenarios/models (called brains) on the go.
- Live view of sensor readings.
The algorithms that command the robots are called brains, and that is where the neural logic (behaviors) is at.
This project presents different approaches to the follow-the-line exercise but using artificial intelligence to complete the circuits. The solutions presented are:
- Using classification networks.
- Using regression networks.
- Using reinforcement learning.
- Solution for real robots.
You can see the project status in the GitHub repository here.